Many buyers assume an infant toy recall means a bad plastic or a toxic finish. This one did not. The CPSC recall notice dated August 27, 2026 (recall number 26-719) states that the glued knot on the elastic cord of the HABA Rainbow Rattle Grasping and Teething Toy can become untied, letting young children reach the small wooden and plastic pieces strung on that cord, with a risk of serious injury or death from choking and ingestion. If you develop or source multi-part teethers, the lesson is not "check your wood." It is that a single assembly method — adhesive plus a knot — was holding the entire small-parts boundary together. Every official fact below is mapped to the component and assembly question it should trigger on your own BOM, and confirmed facts are kept separate from what the notice does not say.
What This Recall Actually Decides for Buyers
- The recalled unit is scoped by item number 2012261001 and batch code 0925, about 2,000 units, sold December 2025 through July 2026 for about $13 — so batch-level traceability, not product-family recall, is what lets you pull the right cartons.
- The hazard mechanism is an untied glued knot on an elastic cord, which means the small-parts compliance boundary sits at a cord-and-knot joint, not at the individual wooden or plastic component.
- Two reports of knots coming untied and no reported injuries: treat this as a low-frequency, high-severity failure mode, and test the joint as a system rather than testing parts one by one.
- Because the remedy is a refund requiring destruction (cut the cord, photo with name and address to [email protected]), importers and retailers carry a reverse-logistics and consumer-instruction cost, not just a stock-pull cost.
What exactly was recalled, and how do I identify affected stock?
The CPSC notice names one product: the HABA Rainbow Rattle Grasping and Teething Toy, item number 2012261001, batch code 0925, about 2,000 units, imported by Habermaass Corporation, Inc. dba HABA USA of Skaneateles, New York, and manufactured in China. Identification is physical, not just by SKU: the batch code appears in black on one of the natural wooden hemisphere pieces, and the item number is on the back side of the backer card above the barcode.
That two-location marking scheme is the practical takeaway. If your own SKU carries the item number only on the carton and the batch only in your ERP, a downstream retailer cannot tell a good unit from a bad one at the shelf. The notice also states the toy was sold at specialty stores nationwide and online at Amazon.com, Target.com, Bellalunatoys.com, Habausa.com and Nordstrom.com — one component failure, six-plus channels, one refund program.
What the notice does not say: it does not claim a mandatory standard was violated for this item, does not give a defect rate, and does not extend to other HABA products, item numbers, or batch codes. Do not repeat those as if they were facts. Federal law prohibits selling products subject to a recall undertaken with CPSC, so the operative action for a buyer is an immediate stock check against both identifiers.
Why does a glued knot on an elastic cord matter more than the parts?
A multi-part grasping toy is a small-parts containment system. Wooden hemispheres, plastic rattling rings, and beads are individually fine because the cord holds them out of reach. The moment the cord's end termination fails, every component on that cord becomes an accessible small part at once.
This is why part-by-part inspection can pass while the assembly fails. Approved incoming-component checks exist for wheels, axles, batteries, fasteners, inserts and printed parts precisely because a finished-toy test may not isolate a supplier defect — the same logic applies to cord, knot and adhesive as purchased inputs. If your supplier buys pre-knotted or pre-glued cord assemblies, that incoming item needs its own check, not a visual glance.
Injection-molded rings and wooden hemispheres may also require secondary assembly, and fit should be checked across the full tolerance stack rather than part by part only. A knot that holds on a loose cord can slip on a tightly strung one, and the recall notice does not specify a cord tension or adhesive type, so you cannot reverse-engineer the exact process from the notice — you can only decide that the joint must be specified and verified.
Mapping Official Recall Facts to Component and Assembly Checks
| Confirmed: hazard mechanism | Glued knot on the elastic cord can become untied, allowing access to small wooden and plastic pieces (CPSC notice, Aug 27, 2026) |
|---|---|
| Buyer check it triggers | Specify cord termination as a controlled process: adhesive type, knot type, cure time, and a mechanical back-up if the design allows it |
| Confirmed: affected identifiers | Item number 2012261001, batch code 0925, about 2,000 units |
| Buyer check it triggers | Require batch marking on the product itself and item number on packaging, so stock can be scoped without pulling the whole family |
| Confirmed: selling window and price | Sold December 2025 through July 2026 for about $13 at specialty stores and named online retailers |
| Buyer check it triggers | Treat an 8-month shelf life and a low price point as recall-exposure multipliers, not as margin comfort |
| Confirmed: reports and injuries | Two reports of knots becoming untied; no injuries reported |
| Buyer check it triggers | Low-frequency, high-severity: sample size on a pull or knot test must be set in the inspection plan, not assumed |
| Confirmed: remedy | Full refund; consumers stop use, cut the elastic cord, photograph the destroyed toy with name and address, email [email protected], then dispose |
| Buyer check it triggers | Budget reverse logistics and a consumer instruction path before a recall, not during one |
| Not stated in the notice | Root cause beyond the untied glued knot; standard violation; defect rate; other SKUs affected |
| Buyer check it triggers | Do not write these as findings. Treat them as open questions for the supplier's process documentation |
Which test documents do I need per market, and why does one certificate never cover it?
Start with the market, then the test scope, then the lab. Not every laboratory or site can perform every toy method, so booking a lab before the destination market and the applicable method are defined usually produces a document that does not match what customs or a retailer's compliance team asks for.
The second rule is that evidence follows the product's structure. A declaration written for an empty plastic shell does not automatically cover the populated assembly, and this is exactly the kind of gap that shows up when a component supplier's paper is reused as if it described the finished toy. A practical baby toy test document checklist by market separates, at minimum: which market, which age grade, which material or component, which test file, and which production lot that file actually covers.
One pattern repeats on the factory floor: a buyer tried to use one certificate for both UKCA and a VOC file on an electric bubble machine. Those are different papers covering different questions. Map market, lot, and which file covers which test. Paperwork that does not match the lot is not paperwork.
For mouthable plastic components, evaluate bite-related release, breakage, sharp points and chemical exposure under the applicable market rule — this is a separate evaluation from a general small-parts assessment, and it applies to the rings in a rattle design as much as to a dedicated teether.
How should I verify a Chinese supplier for a multi-part infant toy?
Ask for the process, not the promise. For a cord-and-knot assembly, the useful documents are the work instruction for knot tying and adhesive application, the cure or dwell time, the incoming inspection record for purchased cord assemblies, and the pull or retention test method with the sample size drawn from the agreed inspection plan on the PO. If a supplier answers with "we test all products," you have learned nothing about the joint.
Second, verify that batch marking is real on the physical unit, not only in the system. The recalled toy was identifiable because the batch code was printed on a wooden hemisphere and the item number sat above the barcode on the backer card. Ask your supplier to show you a production unit or pack with the same fields, and confirm the marking survives normal handling and retail display.
Third, keep the certificate question narrow. A recall is an enforcement or market action, not a substitute for a factory's own test report, and the CPSC notice for this recall does not allege a standard violation. So do not accept "we have a recall-free history" as test evidence, and do not accept a certificate as proof of a specific joint's performance.
Fourth, check how failures are handled. Sorting a failed lot is containment, not root-cause correction. Corrective and preventive action should address the cause of the defect and verify effectiveness. Ask what happens after a failed knot test: is the lot isolated, is the defect class recorded, and what evidence releases it back to production?
Does this recall mean multi-part teething designs are unsafe?
No. It means one specific assembly method failed on one scoped batch, and the notice reports two knot failures with no injuries. The reasonable conclusion is not to abandon multi-part rattles — it is to stop treating the cord termination as a cosmetic step.
The design decision to make is explicit: choose a mechanically locked cord end if the design and age grade allow it, or if adhesive-plus-knot is retained, treat it as a critical-to-quality process with a documented cure, a defined adhesive, and a retention test tied to the lot. Related recall patterns in infant teething products keep circling the same three failure modes — detachable small parts, cords that are too small or too long, and components that fail after freezing — so those three deserve their own line in your design review, whatever the material list says.
One more discipline point for brand developers: do not infer infant suitability from a catalog category, a material name like "natural wood," or product imagery. Require SKU-specific age and intended-use evidence. That is what makes a design review defensible, and it is also what makes a recall scoping exercise take hours instead of weeks.
FAQ
Is the HABA Rainbow Rattle still legal to sell if my stock is batch 0925?
No. Federal law prohibits selling products subject to a recall undertaken with CPSC. Check both identifiers — item number 2012261001 and batch code 0925, printed on the natural wooden hemisphere piece — and pull affected units from shelves and listings immediately.
How many units and what price point were involved?
About 2,000 units, sold from December 2025 through July 2026 for about $13 at specialty stores nationwide and online at Amazon.com, Target.com, Bellalunatoys.com, Habausa.com and Nordstrom.com, per the CPSC notice. A roughly $13 price and an eight-month selling window show how much exposure a low-cost infant SKU can carry.
Were any injuries reported?
No injuries were reported. The firm received two reports of the knots on the elastic cord becoming untied. Do not describe this recall as injury-causing — the hazard is stated as a risk of serious injury or death from choking and ingestion.
What should I require from a supplier making a rattle with an elastic cord?
Require the knot and adhesive process as a specified, documented step: adhesive type, cure time, incoming inspection of purchased cord assemblies, and a retention or pull test with sample size drawn from the inspection plan on the PO. A finished-toy test may not isolate a joint defect, so test the joint as a system.
Can one certificate cover UKCA and a VOC file for the same toy?
No. Those are different documents answering different questions. Decide the market and test scope first, then choose the laboratory, because not every lab or site can perform every toy method, and evidence must follow the product's component and material structure.
What does the CPSC notice not tell me about the root cause?
It states the glued knot can become untied but does not identify an adhesive, cord specification, defect rate, or a mandatory-standard violation, and it does not extend to other item numbers or batches. Treat those as open questions to resolve with your supplier's process documentation, not as findings.
Sources
- HABA USA Recalls Rainbow Rattle Grasping and Teething Toys Due to Risk of Serious Injury or Death from Choking and Inges
- HABA USA Recalls Rainbow Rattle Grasping and Teething Toys Due to Risk of Serious Injury or Death from Choking and Inges
- HABA Rainbow Rattle Grasping and Teething Toy Recall (2026)
- HABA Rainbow Rattle Teething Toy Recalled for Choking Risk: CPSC
Related Products & Categories
Request a Quote or a Factory Audit
If you are developing a multi-part teether, rattle, or grasping toy and want the cord-and-knot joint, batch marking, and market-specific test documents reviewed before tooling is committed, send us the drawing, the target age grade, and the destination market. We will come back with the component and assembly questions we would check first, and a quote or audit scope based on what your design actually contains.